[Date Prev][Date Next] [Chronological] [Thread] [Top]

Re: Synchronization to distant locations over high-latency/low-bandwidth



--On Wednesday, March 15, 2006 3:09 PM +0100 Jan-Piet Mens <jpm@retail-sc.com> wrote:

On Tue Mar 14 2006 at 19:13:35 CET, Quanah Gibson-Mount wrote:

delta-syncrepl would probably be your best bet.  Normal syncrepl has
much  larger bandwidth requirements than slurpd or delta-syncrepl.


That seems to work very nicely (at least with one consumer). I'm unsure
as to the necessity of syncprov-sessionlog parameter on the provider. Is
it still required when using delta-syncrepl?

I don't believe so. I certainly don't use it. I believe its purpose is replaced by the accesslog backend, essentially. ;)

I'm thinking of deploying the exact same configuration to all consumer
slapds. May the replica-id (rid=) in the syncrepl stanza be the same for
all consumers or must it be unique accross all consumers?


The RID can be the same across all consumers.

I use the following across all my replicas:

syncrepl        rid=0
                       provider=ldap://MASTERALIAS.stanford.edu:389
                       bindmethod=sasl
                       saslmech=gssapi
                       realm=stanford.edu
                       searchbase="dc=stanford,dc=edu"
                       logbase="cn=accesslog"

logfilter="(&(objectClass=auditWriteObject)(reqResult=0))"
                       schemachecking=on
                       type=refreshAndPersist
                       retry="60 +"
                       syncdata=accesslog



(The MASTERALIAS bit is a filtered variable at install).

--Quanah


--
Quanah Gibson-Mount
Principal Software Developer
ITS/Shared Application Services
Stanford University
GnuPG Public Key: http://www.stanford.edu/~quanah/pgp.html